'The world needs to slow AI down' Anthropic co-founder tells BBC | BBC News

The rapid advancement of AI systems, now capable of autonomous coordination and potentially dangerous behavior, necessitates a shift from industry-led development to a robust, government-backed regulatory framework. As these systems approach human-level intelligence, the risk of uncontrollable swarms hacking critical infrastructure or financial institutions moves from theoretical to imminent. To mitigate these threats and avoid the loss of public trust, companies must adopt third-party safety audits and adhere to standardized, common-sense security protocols similar to those governing aviation or consumer goods. Furthermore, unchecked AI growth threatens to trigger widespread white-collar unemployment, requiring policymakers to balance technological innovation with social stability. Jack Clark, representing Anthropic, emphasizes that while the potential for scientific and healthcare breakthroughs is immense, the industry must prioritize safety and transparency to prevent catastrophic societal disruption.